Haste scale factors!

[zuxia]

This is true to an extent. Crit still scales at a lower value than haste, even at absurd haste amounts. Since one cannot stack SP gear then the determinate must be made between the secondary stats you will find. If those stats have a 'winner', regardless of amount in said stat, then an average scaling factor can be found over tests by smaller interval (+10, +20, +30 haste) once the basis stats (hit cap and crit soft cap) are reached.

There may be slight variations in the stats based off of having to start and stop casting. Even in those circumstances, a lot of the time you do not have to start completely over, which is what you would be simulating. Your immolate is still ticking on your previous target, your doom is still there, if you had to move before the swap you already cast corruption, etc. The main thing this would show are haste scaling factors for the Northrend Beasts encounter as each boss comes out of the gate, which is not the overarching question of haste scaling for warlocks in general.

That is more the point of this thread I believe.

1) Locks are already getting haste scaling on corruption in 3.3

2) Haste DR is based of a limit of the GCD, research is showing that for locks haste does not actually have a DR in general, jsut at certain points along the stat acquisition path.

Since we have no data on most other classes mechanics in the next expansion, we should not waste time discussing them, especially not all the other classes in a lock specific class forum. This is similar to arguing who will be the president of a given country in 100 years. We don't know, and we can't speculate since we have no data. This thread wasn't initiated as a 'wish list usage for haste for every class thread'; this is a haste scaling factor for warlocks thread. A discussion which you propose would find more banter in the general section of the forums, or at the very least, not in this thread.

You missed a couple of fights in TOC btw...Anub, Faction Champs, and of course Lord J --especially in hard modes --and you if you dont want to inclued Lord J -- thats fine -- Beasts, anub, faction, all make up half the instance -- so I guess that does make a difference --

Not simulating a start over, simulate the phases -- difference -- what I am saying is haste is less valued on shorter fights (i.e. also phases) thus haste scaling factors are incorrect when applied out of theory because no one has taken into the account of what I have stated.

[Modez]

You missed a couple of fights in TOC btw...Anub, Faction Champs, and of course Lord J --especially in hard modes --and you if you dont want to inclued Lord J -- thats fine -- Beasts, anub, faction, all make up half the instance -- so I guess that does make a difference --

Not simulating a start over, simulate the phases -- difference -- what I am saying is haste is less valued on shorter fights (i.e. also phases) thus haste scaling factors are incorrect when applied out of theory because no one has taken into the account of what I have stated.

I do understand what you are saying. I guess the question I should ahve asked, and I did not, is what are you expecting to find? Given the shorter time crit starts coming out ahead? Spirit is now a top stat? You cannot increase spell power, as you can spirit, hit, crit, or haste. Yes, spirit in itself increase spellpower by a factor of 39%, but that is already factored into the sim results. Is your argument that you think: haste will be lower than crit in a shorter fight (let's say 2m average)? It is that haste may be lower than spirit in this same time frame?

I do not see how a stats value will change when we go from one point to another. Yes, mathematically its value may be .05 lower or something given where the boss dies/swap, but haste is much higher than the other stats over a 5m encounter.

That being said, I don't see any harm in proving this one way or another.

EDIT: I am using the same gear as in the second post in this thread. I have not modified any values with the exception of hit to make sure it is setting that to a 0 scale since I am geared for having the 1% spacegoat in my group.





So, crit does matter under a 75s encounter. This is not terribly surprising in hindsight; if you had an encounter that lasted 2.5 seconds and only had an incinerate to cast, you would care most about spell power and crit (due to double damage on crits). This model holds, at least with my gearset, until a fight lasts at least 75s.

That being said, you still have other damage on your previous target as said before, this these numbers still do not fully reflect 'target swapping'.

[Merckx]

So, crit does matter under a 75s encounter. This is not terribly surprising in hindsight; if you had an encounter that lasted 2.5 seconds and only had an incinerate to cast, you would care most about spell power and crit (due to double damage on crits). This model holds, at least with my gearset, until a fight lasts at least 75s.

That being said, you still have other damage on your previous target as said before, this these numbers still do not fully reflect 'target swapping'.

What would you change so that the numbers better reflect target swapping? As it stands, I am already rethinking a lot of my gear choices and am going to start trading haste for crit where it is reasonable. I think Twins is the only encounter in ToC where I am on the same target for 75seconds or more.

[zuxia]

I do understand what you are saying. I guess the question I should ahve asked, and I did not, is what are you expecting to find? Given the shorter time crit starts coming out ahead? Spirit is now a top stat? You cannot increase spell power, as you can spirit, hit, crit, or haste. Yes, spirit in itself increase spellpower by a factor of 39%, but that is already factored into the sim results. Is your argument that you think: haste will be lower than crit in a shorter fight (let's say 2m average)? It is that haste may be lower than spirit in this same time frame?

I do not see how a stats value will change when we go from one point to another. Yes, mathematically its value may be .05 lower or something given where the boss dies/swap, but haste is much higher than the other stats over a 5m encounter.

That being said, I don't see any harm in proving this one way or another.

EDIT: I am using the same gear as in the second post in this thread. I have not modified any values with the exception of hit to make sure it is setting that to a 0 scale since I am geared for having the 1% spacegoat in my group.





So, crit does matter under a 75s encounter. This is not terribly surprising in hindsight; if you had an encounter that lasted 2.5 seconds and only had an incinerate to cast, you would care most about spell power and crit (due to double damage on crits). This model holds, at least with my gearset, until a fight lasts at least 75s.

That being said, you still have other damage on your previous target as said before, this these numbers still do not fully reflect 'target swapping'.

I do agree with you it does not fully reflect target swapping...but it does show the values for haste are incorrect when applied to life -- example you are now dpsn snobolds versus boss -- each mob in faction champs -- anub even

I wanted to point this out since vast majority says crit is worthless always haste -- well once you start considering movement, extra mobs to dps -- the value of haste goes down -- In reality with alot more work I think the "actual value of haste as applied to life" Is alot closer to crit than most think.

[Modez]

I'll try to get some of these various tests run this afternoon or this weekend.

EDIT (not going to link all the pictures - again using 'current gear' with modified hit stat and crit as noted in test):

At 30s, over 2 tests by increasing crit by 0/5% per run, crit is always better than haste by a marginal average of .245.

At 35s, over 2 tests by increasing crit by 0/5% per run, haste is always better than crit by a marginal average of .45.

At 40s, over 2 tests by increasing crit by 0/5% per run, haste is always better than crit by a marginal average of .225.

At 45s, over 2 tests by increasing crit by 0/5% per run, haste is always better than crit by a marginal average of .22.

At 50s, over 2 tests by increasing crit by 0/5% per run, haste is always better than crit by a marginal average of .085.

At 55s, over 2 tests by increasing crit by 0/5% per run, crit is always better than haste by a marginal average of .02.

At 60s, over 2 tests by increasing crit by 0/5% per run, crit is always better than haste by a marginal average of .12.

At 65s, over 2 tests by increasing crit by 0/5% per run, crit is always better than haste by a marginal average of .05.

So there may be specific time intervals where crit will win. These likey occur at the previously discussed lull points in casting where additional haste will not grant you another cast. When one of these periods is reached, crit will have been the better stat as it will have applied that to all of your previous casts. Overall, haste seems to be a more blanket approach to gear for when acquiring new items, as when it is ahead (and it is more frequently) it is ahead by a larger margin than when crit is in the lead.

Summation and averages:

Crit ahead and marginal factor average: (.245 + .02 + .12 + .05)/4 = .108750
Haste ahead and marginal factor average: (.45 + .225 + .22 + .085)/4 = .245

On average, haste is still a better scaling stat, but there may be certain time frames where crit would benefit you more. If we are looking at total iterations of boss fights, haste still wins regardless of time, unless you know every boss time will fall within a specific crit centric time.

[Gerian]

The "haste sweet spot" has existed for quite some time now but seems to have fallen victim to the priority system vs. the rigid rotations we used to have, meaning we've disregarded the actual benefit of fitting one more incinerate in before refreshing immolate/conflagging/etc. The benefit of each point of haste is dependent on its effect on Immolate uptime and Conflagrate/Chaos Bolt delay. As you increase in haste your rotation begins to "desynchronize" until a certain point of inflection at which haste becomes more benficial again.